EVALUATION OF ULTRA-LOW-SPEED JITTER IN ROLLING BALLS

The frictional resistance generated by a ball bearing, when starting from rest, increases to a free rolling value in a finite amount of time instead of experiencing an instantaneous step increase. This is due to the complex ball dynamics involved. A change in the direction of rotation results in a reversal of the bearings dynamic torque characteristics. When a traversing process is continuously repeated, the friction torque of the bearing traces out hysteresis loops. This phenomenon is often described as ball jitter. An apparatus was built to study the ultra-low-speed hysteresis behavior found in jittering ball bearings. This behavior was first modeled by P. R. Dahl (1968). He found torque hysteresis loops to be characterized by the steady rolling friction torque, T(s), and the reversal torque slope, sigma. A theoretical expression for T(s) has been verified in previous research by Lovell et al. (1991), but no expression for sigma has currently been determined at low speeds. Therefore, the experimental testing apparatus was canonically designed to incorporate three balls, rather than a full bearing. This was done in order that the characteristics of sigma could be more readily determined, as many additional friction elements which occur in full bearings were eliminated. The effects of normal load, rotational rate, and lubrication regime on sigma were all investigated using the testing apparatus. The experimental values obtained for the rest slope were coupled with previously determined values of T(s), so that empirical Dahl hysteresis curves could be generated. Excellent correlation was found to exist between the experimental jitter torque data and Dahl predicted dynamic torque behavior.